Michael Kohlhase: AI-Supported Adaptive Course Materials: Didactic Added Value Services based on Semantic LaTeX
**Povzetek. **
Good teachers have a sophisticated model of the knowledge they want to convey and maintain a fine-grained cognitive model of the student's knowledge and preferences. Based on these two models they can generate explanations tailored to individual students or course materials tailored to student cohorts.
We claim that if we can represent knowledge and learner models in a machine-actionable way, we can automate the generation of course materials as active documents: interactive documents that adapt to the student's preferences and current competence.
In the VoLL-KI project at FAU Erlangen https://voll-ki.fau.de we are making this intuition concrete. We are using sTeX https://github.com/slatex/sTeX - a semantic variant of LaTeX - for annotating conventional course materials with semantic information, transform the sTeX into HTML5 which can be instrumented with learner modeling facilities and use that for generated active (interactive and learner-adaptive) course materials.
The talk will informally introduce the underlying ideas, demo an early prototype for "guided tours", and show what instructors have to invest in terms of sTeX markup.